Almost all of the mass distributions that have been proposed to account for the large positive gravity anomalies associated with lunar mascons have assumed single-body sources of a mass excess. In the case of mare fill with a reasonable density contrast (+0.5 g/cm3) with crustal material, a fill thickness of about 16 km for Mare Serenitatis is thus required to account for the observed gravity values at 100-km height. Such a great thickness would require a 16-km-deep hole prior to filling, and such a topographic depression is inconsistent with the depths of the topography of the Mare Nectaris and Mare Oriental basins, which have little fill, and with estimates of mare thicknesses based on buried crater dimensions. A two-body mascon solution, however, requires only about a 2-km thickness of fill and a 12-km rise of a lunar Moho beneath Mare Serenitatis to account for observed gravity anomlaies. The mantle dome results from an uprising of mantle material beneath the mare basin, bringing the impact crater to near isostatic equilibrium. Two kilometers of fill is inferred to have accumulated later, when the crust became rigid enough to sustain the load. Together the fill and the dome account (at about 20% and 80%, respectively) for the magnitude of the observed mascon anomalies. This type of two-body solution can account for greater magnitude mascon gravity anomalies by proportional increases in the fill thickness. The top of the mantle dome or plug is placed at 60-km depth to match observed seismic velocity structure. This mascon structure has an anomalous gravity field that is in agreement with the maximum magnitude of anomalies observed at several heights above Mare Serenitatis. The thickness of fill would be greater if the basin floor had subsided under the load of early fill material.